Thread: The more things change department

The longer run was made using the database program from ChessBase, and yep, 11 days, and still not finished.

One big speed up, was to reconstruct the games as long (very long) strings, moving the PGN tagged info to the very end, and then using system sort -- you can't beat it for string sorting, to sort all those games as strings. Finding dupes is then a piece of cake, and this is where I also believe the database programs fall short.

I don't have a version of BASIC anymore, but it would have also done a fine job with this. It's a DIY program success story, really.

Originally Posted by Adak

The longer run was made using the database program from ChessBase, and yep, 11 days, and still not finished.

One big speed up, was to reconstruct the games as long (very long) strings, moving the PGN tagged info to the very end, and then using system sort -- you can't beat it for string sorting, to sort all those games as strings. Finding dupes is then a piece of cake, and this is where I also believe the database programs fall short.

I don't have a version of BASIC anymore, but it would have also done a fine job with this. It's a DIY program success story, really.

FWIW... the system("sort...") call is actually qsort() which is part of the standard C library... And yes it's probably a little faster to use qsort() than the system call.

But what a nifty success story... 11 days and ticking brought down to 6 minutes... Your friend must have been amazed.

Originally Posted by CommonTater

FWIW... the system("sort...") call is actually qsort() which is part of the standard C library... And yes it's probably a little faster to use qsort() than the system call.

But what a nifty success story... 11 days and ticking brought down to 6 minutes... Your friend must have been amazed.

The Windows system sort appears to be a multi-merge sort (if I understand their cryptic description correctly), that balances out the merging passes. It has guaranteed resources and low level micro code, that our sorting programs can only dream about.

My optimized Quicksort's leave qsort() in the dust, and sysem(sort...), leaves my optimized Quicksorts, far behind.

I ran across this first, in Windows2000, and it's been true ever since ( Vista not checked). I've tested it with millions of Sudoku grids, and billions of chess positions (in FEN strings), and there is simply no comparison of their sorting times. Try it and see.

I haven't tried it with int's or floats, because I like my Quicksort!

One change for Windows 7 system sorting is that Windows now sees any digit in the string, as a NUMBER, not a char. That's a helluva change, imo, and when you make up the maximum size of the record to sort for system sort syntax, you have to add extra size, to accommodate that change.

Non programmers are always amazed by the difference. "How could this little console program beat my $$$ database program? It's an epiphany for them.